Search papers, labs, and topics across Lattice.
3
0
8
Dataset distillation gets a boost on long-tailed data with CSDM, which uses spectral distribution matching to prioritize realism in tail classes and achieves a 14% improvement over SOTA methods.
Current vision-language models are surprisingly bad at surgical safety reasoning, failing to integrate phase information to identify safe operative zones, but a new RLHF-tuned model closes the gap.
Achieve SOTA joint audio-video generation with JavisDiT++ using just 1M public training examples, rivaling performance of models trained on proprietary datasets.
